PF1 Crystal structure of IcaR, a repressor of the TetR family implicated in biofilm formation in Staphylococcus epidermidis

Wen-Yih Jeng*, 1, 2, Tzu-Ping Ko*, 1, 2Rey-Ting Guo1, 2, 3Chien-Liang Liu1, 4Chia-I Liu1, 2, 3Hui-Lin Shr1, 2Andrew H.-J. Wang1, 2, 3

*These authors contributed equally to this work. 1Institute of Biological Chemistry, 2Core Facility for Protein Crystallography, Academia Sinica, Taipei 115, Taiwan; 3Institute of Biochemical Sciences, National Taiwan University, Taipei 106, Taiwan; 4Structural Biology Program, Institute of Biochemistry and Molecular Biology, National Yang-Ming University, Taipei 112, Taiwan


Expression of the gene cluster icaADBC is necessary for biofilm production in S. epidermidis. The ica operon is negatively controlled by the repressor IcaR. Here, the crystal structure of IcaR was determined by multiple anomalous diffraction methods, using a seleno-methionine derivative. It revealed a homodimer comprising entirely of α-helices, typical of the tetracycline repressor protein family for gene regulations. The N-terminal domain contains a conserved helix-turn-helix DNA-binding motif. Conformational variation indicates flexibility in this region. The C-terminal domain shows a complementary surface charge distribution about the dyad axis, ideal for efficient and specific dimer formation. The results of electrophoretic mobility shift assay suggest that a 28-base-pairs core segment of the ica operator is implicated in cooperative binding of two IcaR dimers on opposite sides of the DNA. Computer modeling based on the known DNA-complex structure of QacR showed that direct protein-DNA interactions are mostly conserved, but with slight variations for recognizing the different sequence. By interfering with the binding of IcaR to DNA, gentamicin and other antibiotics may elicit biofilm production in S. epidermidis, as a defense mechanism.




Long-Range Cooperative Interactions Modulate Dimerization in S. cerevisiae Geranylgeranyl Pyrophosphate Synthase

Chia-Hsiang Lo1, Tao-Hsin Chang2, Hui-Chuan Chang2, and Po-Huang Liang1,2

1Institute of Biochemical Sciences, National Taiwan University, Taipei 106, Taiwan
2Institute of Biological Chemistry, Academia Sinica, Taipei 115, Taiwan


Geranylgeranyl pyrophosphate synthase (GGPPs) catalyzes the condensation reaction of farnesyl pyrophosphate (FPP) with isopentenyl pyrophosphate (IPP) to generate C20 geranylgeranyl pyrophosphate, which is a precursor for carotenoids, chlorophylls, geranylgeranylated proteins, and archaeal ether linked lipid. The structure is composed entirely of fifteen α-helices joined by connecting loops and is arranged with α-helices around a large central cavity. Distinct from other known structures of trans-prenyltransferases, the N-terminal 17amino acids (9-amino acid helix A and the following loop) of this GGPPs protrude from the helix core into the other subunit and contribute to the tight dimer formation. Deletion of the first 9 or 17 amino acids caused the dissociation of dimer into monomer and the del(1-17) mutant showed abolished enzyme activity. Here, we further demonstrate that deletion of the first 7 residues in N-terminus is sufficient to disrupt the dimer into the mixture of monomer and dimer and deletion of 8 residues causes complete monomer. To identify a single amino acid important for dimerization, E7G, L8G, I9G with each of the amino acids at position 7, 8, and 9 replaced with Gly were characterized using analytical ultracentrifuge(AUC). E7G is a dimer, L8G is a mixture and I9G is a major monomer, indicating that Leu8 and Ile9 are essential in dimerization. This represents a novel case in which the amino acids at N-terminus are far away from the dimer interface (26Å), but significantly contribute to the dimer formation.



PF3 Structural Characterization of the ApoE C-terminal Domain by NMR Spectroscopy

Pei-Hsin Liao1,2, Chi-Jen Lo1,2, Hui-Ling Wang3, Hsien-Bin Huang3, Yi-Chen Chen4, Chi-Yuan Chou5, Gu-Gang Chang2,5, Chinpan Chen6, Ta-Hsien Lin1,2,7

1Institute of Biochemistry and Molecular biology, 2Structural Biology Program, National Yang-Ming University, Taipei 112, Taiwan
3Institute of Molecular Biology, National Chung Cheng University, Chia-Yi 621, Taiwan
4Department and Institute of Medical Biotechnology, Tzu Chi University, Hualien 970, Taiwan
5Faculty of Life Sciences and Institute of Genome Sciences, National Yang-Ming University, Taipei 112, Taiwan
6Institute of Biomedical Sciences, Academia Sinica, Taipei 115, Taiwan
7Department of Medical Research & Education, Taipei Veterans General Hospital, Taipei 112, Taiwan


Human apolipoprotein E (apoE) is a 299-residue polymorphic protein (Mr = 34 kDa). It is composed of two independently folded domains that are separated by a hinge region. The 22-kDa N-terminal domain (residues 1-191) and the 10-kDa C-terminal domain (residues 216-299) of apoE are responsible for receptor binding and lipid binding, respectively. The APOE gene has three major alleles, ε2, ε3 andε4, which encode three major isoforms, apoE2, apoE3, and apoE4, respectively. The three isoforms differ from one another only at residue 112 and 158, but have marked differences in their biological functions. ApoE has been known to play a key role in the transport of plasma cholesterol and lipoprotein metabolism. The lipid- and receptor-binding abilities of apoE are isoform-specific. Besides, apoE is also highly associated with late-onset familial and sporadic Alzheimer’s disease (AD). The alleleε4 is a risk factor for Alzheimer’s disease. The structural characteristics of apoE isoforms may play important roles in their biological functions.
The structure-functional relationship of apoE isoforms may be obtained from their three-dimensional structures. The three-dimensional structures of the N-terminal domain of apoE isoforms have been solved by x-ray crystallography, but the three-dimensional structures of the C-terminal domain of apoE and intact apoE isoforms still haven’t been solved. In this study we have applied NMR spectroscopy, CD spectroscopy and AUC to characterize the structure of apoE (195-299) in aqueous solution. The results suggested that apoE C-terminal domain form a trimer in aqueous solution with high helix content.




Dimerization is important for the GTPase activity of chloroplast translocon components atToc33 and psToc159

Yi-Hung Yeh1,2, Muppuru M. Kesavulu1, Hsou-min Li1, Shu-Zon Wu1, Yuh-Ju Sun2, Emadeldin H. E Konozy1 and Chwan-Deng Hsiao1

1Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan 115, ROC; 2Institute of Bioinformatics and Structural Biology, National Tsing Hua University, Hsinchu, Taiwan 300, ROC


Arabidopsis Toc33 (atToc33) is a GTPase and a member of the Toc (translocon at the outer-envelope membrane of chloroplasts) complex that associates with precursor proteins during protein import into chloroplasts. By inference from the crystal structure of psToc34, a homologue in pea, the arginine at residue 130 (Arg130) has been implicated in the formation of the atToc33 dimer and in inter-molecular GTPase activation within the dimer. Here we report the crystal structure at 3.2 Å resolution of an atToc33 mutant, atToc33(R130A), in which Arg130 was mutated to alanine. Both in solution and in crystals, atToc33(R130A) was present in its monomeric form. In contrast, both wild-type atToc33 and another pea Toc GTPase homologue, pea Toc159 (psToc159), were able to form dimers in solution. Dimeric atToc33 and psToc159 had significantly higher GTPase activity than monomeric atToc33, psToc159 and atToc33(R130A). Molecular modeling using the structures of psToc34 and atToc33(R130A) suggests that, in an architectural dimer of atToc33, Arg130 from one monomer interacts with the β-phosphate of GDP and several other amino acids of the other monomer. These results indicate that Arg130 is critical for dimer formation, which is itself important for GTPase activity. Activation of GTPase activity by dimer formation is likely to be a critical regulatory step in protein import into chloroplasts.




ATP-driven Motions of 70-kDa Heat Shock Proteins (Hsp70s): Insights into Structural Dynamics of the Hsp70 Power Stroke

Yi-Wei Chang1,2, Chung Wang1, and Chwan-Deng Hsiao1

1Institute of Molecular Biology, Academia Sinica, Taipei, 11529, Taiwan;
2Institute of Bioinformatics and Structural Biology, National Tsing Hua University, Hsinchu, 300, Taiwan


The 70-kDa heat shock proteins (Hsp70s) are interspecific highly conserved molecular chaperones which participate in many crucial cellular functions, such as promoting the correct folding of nascent or stress-denatured polypeptides, assisting the protein translocation across membranes, and helping the assembly/disassembly of protein complexes. Under particular control by the J-domain ATPase activating proteins and nucleotide exchange factors (NEFs), Hsp70s process their substrates in ATP-driven cycles. Previous studies clearly defined that Hsp70 protein can be distributed into two major functional domains, the NH2-terminal nucleotide-binding domain (NBD) and the COOH-terminal substrate-binding domain (SBD). To date, the ATP dependent allosterically regulating mechanism between these two domains has been long explored. However, precise structural understanding of the interdomain communication during ATPase cycle is still limited. In this report, we have determined the X-ray structures of ATP form intact Hsp70 chaperone proteins in various species, including the DnaK in eubacterium Geobacillus kaustophilus HTA426 and the 70-kDa heat shock cognate protein (Hsc70) in mammalian Rattus norvegicus. Together with the result of detecting the solvent accessibility of a Trp reporter on the domain-domain interface, the structures reveal typical domain releasing behaviors of Hsp70s upon ATP binding and hydrolysis. This is in good agreement with previous studies that the hydrolysis of ATP may drives large domain movement of SBD related to NBD and provides power to allow Hsp70s serving as motors in processes such as protein translocation trough transport channels. Furthermore, the insertion of the hydrophobic linker region of gkDnaK into another crystallographic symmetry molecule’s substrate binding pocket suggests a characteristic cooperative mechanism of Hsp70s. These important findings demonstrate exciting insights into structural dynamics of Hsp70 family proteins during chaperone cycles.




Structural basis of PhoP-activated PmrD protein that connects PhoP/PhoQ and PmrA/PmrB two component systems

Shih-Chi Luo 1.2, Yuan-Chou Lou2, and Chinpan Chen2

1 Taiwan International Graduate Program, Academia Sinica, Nan-Kang, Taipei, 115, Taiwan
2Institute of Biomedical Science, Academia Sinica, Nan-Kang, Taipei, 115, Taiwan


Bacterial have many two-component signal-transduction systems (TCSs) that response to specific environmental signals by altering the phosphorylated state of a response regulator (RR). It was found recently that PmrD can mediate the activation of a PmrA-PmrB two-component system by another PhoP-PhoQ two-component system that responds to the magnesium levels in the environment in Salmonella. PmrD is the first example of a small basic protein that enables a TCS to response to the signal govering a different TCS by protecting the phosphorylated state of RR. Here we report the solution structure of PmrD from Klebsiella pneumoniae, which shows a beta-barrel like protein consisting of a six-stranded antiparallel beta-sheet, on the basis of 1033 restraints. The ensemble of 20 NMR structures was well defined, with average root-mean-square deviations of 0.46 ± 0.11 Å for the backbone atoms and 1.35 ± 0.26 Å for heavy atoms. Functional analyses of PmrD showed that it confers polymyxin B resistance activity in Klebsiella pneumoniae and retains higher levels of PmrA phosphorylation in the same way as in Salmonella. Besides, very little is known about the structural aspects of interactions between PmrD and phosphorylated PmrA. We address this issue by using NMR chemical shift mapping to identify the binding interface. Compared to the 2D HSQC spectrum of the free PmrD, we found that the N-terminal region (β1 to β3) and the C-terminal loop of PmrD posess the most perturbed chemical shift environments, which indicates that they may be the binding sites for the interaction with phopho-PmrA.




RNA degradation by Escherichia coli polynucleotide phosphorylase

Zhonghao Shi1,2 , Wei-Jen Yang 2 , Kin-Fu Chak 1 and Hanna S. Yuan2

1Institute of Biochemistry and Molecular Biology, National Yang-Ming University, 2Institute of Molecular Biology, Academia Sinica


RNA degradation is an important biological process in regulating mRNA levels in cells. In eukaryotes, a protein complex, exosome, plays a major role for mRNA degradation in cytoplasm. The bacterial trimeric polynucleotide phosphorylase (PNPase) shares similar sequence and domain organization and bears similar function in mRNA degradation to that of eukaryotic exosome. We determined the crystal structure of E. coli PNPase at a resolution 2.7 Å. PNPase, containing two RNase PH domains and a S1/KH domain, forms a trimer. The six RNase PH domains of the trimeric PNPase fold into a ring-like structure containing a central channel for RNA binding and degradation. This ring-like architecture of bacterial PNPase is similar to those of archeal and human exosomes, supporting the hypothesis that a common mechanism for RNA degradation is used in all kingdoms of life. We also found that the wild-type PNPase produces a final end product of 4-nucleotide RNA. However, a R106A mutant and a S1/KH domain truncated PNPase produce different RNA end products, indicating that R106 located inside the channel and S1/KH domain are likely involved in RNA recognition.




Structural and kinetic characterization of Arctic β-amyloid peptide by NMR spectroscopy

Chih Ching Wang1,2, Yi-Chen Chen3, Chia-li Su4, Hun-Gen Chang4, Zi-Fan Wang4, Hsien-Bin Huang4 and Ta-Hsien Lin1,2,5

1Institute of Biochemistry and Molecular Biology, 2Structural Biology Program National Yang-Ming University, Taipei 112, Taiwan
3Department and Institute of Medical Biotechnology, Tzu Chi University, Hualien 970, Taiwan
4Institute of Molecular Biology, National Chung Cheng University, Chia-Yi 621, Taiwan
5Department of Medical Research & Education, Taipei Veterans General Hospital, Taipei 112, Taiwan


Aggregation of β-amyloid peptide (Aβ) has been thought to be an early event in the pathogenesis of Alzheimer’s disease. In vitro studies have shown that the genetically mutated Aβ (Arctic mutant) had a higher propensity to form protofibrils than wild-type Aβ. The molecular mechanism of Arctic mutant prone to aggregation is still not entirely understood. By applying heteronuclear multidimensional NMR spectroscopy we have determined the structure of Arctic Aβ mutant (Aβ40(E22G)) in aqueous sodium dodecyl sulfate (SDS) micelles. Comparisons of the structures of Aβ40(E22G) and wild-type Aβ40 reveal that Arctic mutation leads to α-helix to random coil conformational changes in the Q15KLVFFAEDV24 region of Aβ40(2). The kinetic aggregation profiles of Aβ40(E22G) and wild-type Aβ40 obtained by using NMR spectroscopy suggest that both peptides undergo nucleation-dependent polymerization in aqueous SDS micelles. Besides, the lengths of lag phase and growth phase of Aβ40(E22G) are shorter than those of wild-type Aβ40, indicating that Aβ40(E22G) is more prone to aggregation than wild-type Aβ40. The results of structural and kinetic studies suggest that the less helix content for Aβ40(E22G) may result in increasing the tendency of α-helix/β-strand conversions, which in turn accelerate the rate of oligomerization of Arctic mutant. These results also provide a structural basis towards understanding the role of Arctic mutation in accelerating Aβ aggregation.




The dodecameric structure of pyridoxal 5'-phosphate-containing L-aspartate 4-decarboxylase

Hui-Ju Chen1,2, Tsu-Ping Ko1, Nai-Chen Wang3, Chia-Yin Lee3, Andrew H.-J. Wang1,2

1 Institute of Biological Chemistry, Academia Sinica, Taipei 115, Taiwan. 2 Institute of Biochemical Sciences, National Taiwan University, Taipei 106, Taiwan. 3 Graduate Institute of Agricultural Chemistry, National Taiwan University, Taipei 106, Taiwan.


L-Aspartate 4-decarboxylase (ASD ; E.C. catalyses the removal of the beta carboxyl group from L-aspartate with complete retention of the asymmetric centre to produce L-alanine. It is an important enzyme to produce L-alanin and D-aspartate in prokaryote. The recombinant ASD from Alcaligenes faecalis CCRC 11585 (asdA) and Pseudomonas sp. ATCC 19121 (asdP) of about 533 amino acids, with a calculated molecular weight of 59584 and 62123 Da and a pI of 5.14 and 6.75, respectively. Base on those recombinant sequences, the asd gene was overproduced in Escherichia Coli system and purified to homogeneity. Crystallographic studies of asdA and asdP were initiated in order to study the relationship of the assembly structure and the mechanism of D-amino acid produced. By the hang-in vapor-diffusion method, the asdA protein crystallized using 0.1M Tris-HCl pH 7.0-8.5, 0.1M LiSO4, 20% PEG4000 at 277K for 30 days to 10 month.
Preliminary X-ray analysis of asdA showed that the crystal belongs to space group C2221, with unit-cell parameters a=150.7Å, b=217.2Å, c=209.4Å and six molecule per asymmetric unit. A comparable 2.5Å data set was measured at 180K using a synchrotron of NSRRC source and Raxis IV imaging plate area detector (96.6% completeness, Rmerge = 10.7%). The asdP protein was crystallized on sitting plate with 20% PEG400, and 0.1M Tris-HCl pH8.5, 0.2M LiSO4 system. The space group of asdP were analyzed to F432 (unit cell a=b=c 298Å ) . The data of methyl HgCl derivated asdP had solved for the phase problem. Molecular replacements are currently in progress in order to determine the detailed three-dimensional structure of asdA.
The analytic ultracentrifugation reveal the ASD protein were assemble to dodectamer conformation on native form. Also, the basic structure of ASD depends on dimmer conformation that was a foundation of functional structure. The six dimmer structures contributed to a dodecatamer and the N-terminal and C-terminal amino acids were important on the enzymatic kinetic of aspartate decarboxylation.




Crystal Structure of human Tudor-SN and Implication of its Roles in RNA Interference

Chia-Lung Li (李家隆) and Hanna S. Yuan(袁小琀)

Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan (中央研究院 分子生物研究所)


The human Tudor-SN (also called p100) was first identified as a co-activator of Epstein-Barr virus nuclear antigen (EBNA-2). Tudor-SN has been reported to associate with several transcriptional factors, including TFIIE, c-Myb and STAT6, to enhance the transcriptional activity. Tudor-SN has also been found as one of the component proteins in the RNA interference silencing complex (RISC), and therefore it is likely involved in RNA silencing. Recently Tudor-SN has been shown to bind to hyper-edited double-stranded microRNA and promotes its cleavage. Tudor-SN contains four tandem repeats of staphylococcal-nuclease-like domains (SN domain) at N-terminus and a Tudor domain linked with a partial SN domain at C terminus. Although extensive functional studies have been carried out, the structure and domain arrangement of Tudor-SN is unknown. The structural based analysis of Tudor-SN is thus imperious for better understanding of the role of this protein in RNA interference and transcription activation. We crystallized a stable 70-kD truncated form of human Tudor-SN and solved the structure by MAD method at a resolution of 1.9 Å. X-ray diffraction data were recorded at the Taiwan beamline BL-12B in SPring-8. The overall Tudor-SN structure contains two intact SN domains (SN3 and SN4), one Tudor domain and a C-terminal partial SN domain (SN5). Superposition of Tudor-SN with staphylococcal nuclease shows that some of the important active site residues of SN3 and SN4 domains are missing, indicating that some of the SN domains, including SN4, may not be active in RNA cleavage. Structural superposition of the OB fold in the SN domain of Tudor-SN with the OB-fold structure of Sac7d/DNA complex reveals a substrate binding cleft located between SN3 and SN4 interface. We suggest that SN3 domain is likely responsible for RNA cleavage and SN4 domain is likely involved in substrate binding but not in cleavage. These results are consistent with the activity assays for different truncated forms of Tudor-SN in RNA cleavage. In conclusion, the structural analysis of Tudor-SN suggests a coordination between two SN domains in RNA binding and cleavage and provides some clues for further functional investigations.




Carbohydrate-binding of the starch binding domain of Rhizopus oryzae glucoamylase in complex with β-cyclodextrin and maltoheptaose

Jung-Yu Tung , Yen-Yi Liu, Wei-I Chou, Fang-Yu Chang Dah-Tsyr Chang and Yuh-Ju Sun

Department of Life Science, National Tsing Hua University, Hsinchu


Glucoamylase hydrolyses starch and polysaccharides to β-D-glucose. Rhizopus oryzae glucoamylase (RoGA) consists of two functional domains, an N-terminal starch binding domain (SBD) and a C-terminal catalytic domain. The two domains are connected by an O-glycosylated linker. The crystal structures of the SBD in complex with a cyclic carbohydrate, β-cyclodextrin (β-CD), and a linear carbohydrate, maltoheptaose (G7), were determined. The overall structures of the SBD complexes belong to a β-sandwich fold with an immunoglobulin-like structure and two carbohydrate-binding sites were observed. Site I is created by several conserved aromatic residues, Trp47, Tyr83, and Tyr94, to form a broad, flat, and firm hydrophobic surface. In addition to the hydrophobic interaction, numerous asparagine residues are involved in the hydrophilic interactions to carbohydrate. Site II is built up by Tyr32 and Phe58 to produce a protruded and narrow binding environment. Site I undergoes a bigger conformational change than site II upon the carbohydrate binding.




Binding of oligonucleotides to SARS-CoV nucleocapsid protein dimerization domain: A nuclear magnetic resonance study

Chung-ke Chang, Yen-lan Hsu, Yuan-hsiang Chang, Tai-huang Huang

Institute of Biomedical Sciences, Academia Sinica No. 128, Academia Road Section 2, Nan-Kang, Taipei, 115, Taiwan


The SARS-CoV nucleocapsid protein (NP) comprises two structural domains and encapsidates viral genomic RNA. The dimerization domain (DD) contains a positively charged region capable of binding to DNA and RNA. Using nuclear magnetic resonance techniques, we identified the binding sites and estimated the dissociation constants of DD towards a variety of single-stranded DNA oligonucleotides. Structural implications in relation to the binding mechanism are discussed.




NMR Studies of Virulence-associated Proteins and Small Conserved Hypothetical Proteins in Klebsiella Pneumoniae

Kuo-Wei Hung 1, Chun-Chia Cheng1, Yi-Chao Lin1, Jia-Huei Chen1, Tsan-Hung Yu1, Pei-Ju Fan2, Chi-Fon Chang2, Shih-Feng Tsai3 and Tai-Huang Huang1,2,*

1Inst. Biomed. Sci., 2Genomic Research Center, Academia Sinica, Taipei, Taiwan, ROC
3Div. Molecular & Genomic Medicine, National Health Research Institute, Zhunan, Miaoli, Taiwan, ROC


Klebsiella pneumoniae is an enteric gram-negative bacillus causing hospital-acquired infections of immunocompromised patients. Mortality rates as high as 10% for primary liver abscess, and among them, 30~ 40% for those with metastatic meningitis have been reported. In the past two decades, primary K. pneumoniae liver abscess and its septic metastatic complications have emerged as one of the most common community-acquired bacterial diseases in Taiwan. We have selected 34 target proteins in K. pneumoniae for the structure determination by NMR strategies. 14 target proteins possessing high solubility have been isotope-labeled for the investigation of structure properties by 15N-1H HSQC screening. Using a variety of triple resonance NMR experiments, complete assignments of 1H, 15N and 13C resonances have been accomplished for 3 target proteins including KP5110, KP1966 and KPP136C103 (BioMagResBank accession number 7062, 7204 and XXXX, respectively). In addition, the near complete backbone resonances of 6 target proteins KP0001, KP5111N267, KP5112, KP3708, KP1972 and KP2453 have been achieved. The three-dimensional structures of KP5110, KP1966 and KPP136C103 have been calculated using distance-geometry followed by simulated annealing techniques with ARIA-CNS and CYANA (PDB ID 2GCX, 2DUW and XXXX, respectively). Iron acquisition system is one of the most important pathogenic factors of Klebsiella pneumoniae. Feo (ferrous iron transport) proteins including FeoA (KP5110), FeoB (KP5111) and FeoC (KP5112) play a specific role for acquiring Fe2+ from surroundings or their hosts. In the present studies, we report the first three-dimensional structure of FeoA and characterize the binding interactions among bio-molecules from Feo system.




Solution structure of Kazal-type Serine Protease Inhibitor 2

Ting Chen, Tian-Ren Lee and Ping-Chiang Lyu

Institute of Bioinformatics and Structural Biology, National Tsing Hua University, Hsinchu,Taiwan


Kazal-type serine proteinase inhibitor 2 (SPINK2), a human seminal plasma proteinaceous inhibitor, is defined by a conserved Kazal domain. A typical Kazal domain contains six cysteine residues that form disulfide bridges stabilizing conformation of the domain. The SPINK2 gene was reported to show the greatest differential expression levels in primary cutaneous follicle center cell lymphomas, hematopoietic stem progenitor cells and human testis cell involved in fertility. Here, we employed an effective strategy to produce and purify SPINK2 protein from Escherichia coli. A DNA fragment carrying SPINK2 gene was amplified by a PCR-based two-step DNA synthesis method, and the recombinant SPINK2 (rSPINK2) was expressed by E. coli. The rSPINK2 was successfully purified by affinity chromatography and high performance liquid chromatography (HPLC). The structural characteristics were checked by biophysical and biochemical approaches. By using trypsin activity assay, the result showed that rSPINK2 has an ability to inhibit trypsin. The result of protein sequence alignment with previously reported trypsin inhibitors revealed that Arg24 is the residue at P1 site and which was also confirmed by site-directed mutagenesis experiments. In addition, three-dimensional structure of SPINK2 was determined by nuclear magnetic resonance spectroscopy. Although SPINK2 has < 45% sequence identity with other trypsin inhibitors with known structure, it was found to contain an α-helix and one triple-stranded anti-parallel β-sheet which similar with others. Additionally, using computational docking experiments to analyze the interaction between trypsin and SPINK2, we identified the possible binding site, which was consistent with the results of sequence alignment and site-directed mutagenesis.




The metabolomic and functional genomic studies of protein stability

Ta-Peng Chiu1, Jia-Hua Wu1, Ching-Yao Chang1, Nai-Wan Hsiao2

1 Asia University, Taichung 413, Taiwan; 2National Changhua University of Education, Changhua 500, Taiwan e-mail:


Proteomics and genomics are primary subjects to solve out the sequences and structures, but not all proteins can solve out their structures smoothly. We analyse through the integration of the databases(TargetDB、KEGG) that can find out more meaningful information, and then we combine the algorithm of predict the protein temperature stability that developed in the past in the laboratory : Temperature Index, and get Instability Index that predict the protein stability in protein's half-life. Using their primary sequences to analyse these proteins which are been solving structures and separate the proteins that can not solve out the structures from that solve out into two kinds of structures to do analysis, and then go to predict the chance of solving out the structures of proteins that will begin to solve of the structures soon. We make two algorithms to be on-line tool, stabilize under tool of factor cooperate through two focus each other in different protein. This tool leaves a more accurate suggestion on users, and web lie between Taxi on-line tool development the stabilize under tool of factor cooperate through two focus each other in different protein. The users can use the analysis of making an amendment in protein project.




A ligand-based molecular modeling studies on a series of tyrosinase Inhibitors using 3D-QSAR CoMFA and CoMSIA

You-Tzung Juang1 , Keng-Chang Tsai , Ching-Yao Chang1 , Nai–Wan Hsiao2

1Asia Universty , Taichung, 413, Taiwan; 2National Changhua University of Education, Changhua, 500, Taiwan


Tyrosinase (EC is a copper-containing enzyme and plays an important role in melanin biosynthesis, which is involved in determining the color of mammalian skin and hair. The state-of-the-art three-dimensional quantitative structure-activity relationship (3D-QSAR) studies using comparative molecule field analysis (CoMFA) and comparative molecule similarity indices analysis (CoMSIA) techniques were implemented to analyze the binding affinity of two sets of N-Benzylbenzamides and Tetraketones inhibitors of tyrosinase. The optimal CoMFA and CoMSIA models obtained from the training set were all statistically significant. The contour maps of the best CoMFA and CoMSIA models were used to identify the structural features relevant to the biological activity in these series of analogs. The result suggests that the 3D-QSAR models built in this paper can be used to guide the development of novel inhibitors of tyrosinase.




Structural and functional studies of the 6-phosphogluconate dehydrogenase associated with pathogenesis of Klebsiella pneumoniae

Ying-Yin Chen(陳盈頴)1, Tzu-Ping Ko (柯子平)2, Li-ping Lo(羅麗萍)2, Chun-Hung Lin(林俊宏)2, Andrew HJ Wang(王惠鈞)2

1Institute of Biochemical Sciences, National Taiwan University, Taipei, Taiwan; 2Institute of Biological Chemistry, Academia Sinica, Taipei, Taiwan


The Gram-negative bacterium, Klebsiella pneumoniae, is an opportunistic pathogen that is commonly isolated from urinary tract infections, nasocomial pneumonia, surgical wounds and blood stream infections. The enzymes directly related to its virulence are validated as novel targets for the understanding of the pathogenicity of K. pneumoniae, development of new antibiotics, as well as decoding of the structure and functional relationship of innate immunity. The 6-phosphogluconate dehydrogenase (6PGDH) has been determined and annotated as a metabolic enzyme of the pentose phosphate (PPP) pathway. Recent finding demonstrates that 6PGDH is one of cell wall lectin proteins and elicits a protective immune response in mice. Our study identifies the structures of 6PGDH from E. coli and K. pneumoniae and predicts the carbohydrates binding region for its adhesion function. The resulting high resolution three-dimensional structures will provide valuable information to design potent enzyme inhibitors as potential new drug candidates.




Crystal structure of the human FOXO3a-DBD/DNA complex suggests the effects of post-translational modification

Kuang-Lei Tsai 1,2, Yuh-Ju Sun 2, Cheng-Yang Huang 1, Jer-Yen Yang 3, Mien-Chie Hung 3 and Chwan-Deng Hsiao 1

1 Institute of Molecular Biology, Academia Sinica, Taipei, 115, Taiwan;
2 Institute of Bioinformatics and Structural Biology, National Tsing Hua University, Hsinchu, 300, Taiwan;
3 Deaprtment of Molecular and Cellular Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030, USA


FOXO3a is a transcription factor of the FOXO family. The FOXO proteins participate in multiple signaling pathways, and their transcriptional activity is regulated by several post-translational mechanisms, including phosphorylation, acetylation, and ubiquitination. Because these post-translational modification sites are located within the C-terminal basic region of the FOXO DNA-binding domain, it is possible that these post-translational modifications could alter the DNA-binding characteristics. To understand how FOXO-mediated transcriptional activity, we report here the 2.7 Å crystal structure of the DNA-binding domain of FOXO3a (FOXO3a-DBD) bound to a 13-bp DNA duplex containing a FOXO consensus binding sequence (GTAAACA). Based on a unique structural feature in the C-terminal region and results from biochemical and mutational studies, our studies may explain how FOXO-DBD C-terminal phosphorylation by protein kinase B (PKB) or acetylation by cAMP-response element binding protein (CBP) can attenuate the DNA-binding activity and thereby reduce transcriptional activity of FOXO proteins. In addition, we demonstrate that the methyl groups of specific thymine bases within the consensus sequence are important for FOXO3a-DBD recognition of the consensus binding site.




Site-directed mutagenesis on the heme axial-ligands of cytochrome b559 in photosystem II by using cyanobacteria Synechocystis PCC 6803

Chung-Hsien Hung, Jing-Yueh Huang, Yi-Fang Chiu, Hsiu-An Chu.

Institute of Plant and Microbial Biology, Academia Sinica No. 128, Academia Road Section 2, Nan-Kang, Taipei, 115, Taiwan


Cytochrome (cyt) b559 has been proposed to play an important role in the cyclic electron flow processes that protect photosystem II (PSII) from light-induced damage during photoinhibitory conditions. However, the exact role(s) of cyt b559 in the cyclic electron transfer pathway(s) in PSII remains unclear. To study the exact role(s) of cyt b559, we have constructed a series of site-directed mutants, each carrying a single amino acid substitution of one of the heme axial-ligands, in the cyanobacterium Synechocystis sp. PCC6803. In these mutants, His-22 of the α or the β subunit of cyt b559 was replaced with either Met, Glu, Tyr, Lys, Arg, Cys or Gln. On the basis of oxygen-evolution and chlorophyll a fluorescence measurements, we found that, among all mutants that were constructed, only the H22Kα mutant grew photoautotrophically, and accumulated stable PSII reaction centers (.81% compared to wild-type cells). In addition, we isolated one pseudorevertant of the H22Yβ mutant that regained the ability to grow photoautotrophically and to assemble stable PSII reaction centers (.79% compared to wild-type cells). On the basis of 77 K fluorescence emission measurements, we found that energy transfer from the phycobilisomes to PSII reaction centers was uncoupled in those cyt b559 mutants that assembled little or no stable PSII. Furthermore, on the basis of immunoblot analyses, we found that in thylakoid membranes of cyt b559 mutants that assembled little or no PSII, the amounts of the D1, D2, cyt b559α and β polypeptides were very low or undetectable but their CP47 and PsaC polypeptides were accumulated to the wild-type level. We also found that the amounts of cyt b559β polypeptide were significantly increased (larger than two folds) in thylakoid membranes of cyt b559 H22YβPS+ mutant cells. We suspected that the increase in the amounts of cyt b559 H22YβPS+ mutant polypeptides in thylakoid membranes might facilitate the assembly of functional PSII in cyt b559 H22YβPS+ mutant cells. Moreover, we found that isolated His-tagged PSII particles from H22Kα mutant cells gave rise to redoxinduced optical absorption difference spectra of cyt b559. Therefore, our results concluded that significant fractions of H22Kα mutant PSII particles retained the heme of cyt b559. Finally, this work is the first report of cyt b559 mutants having substitutions of an axial heme-ligands that retain the ability to grow photoautotrophically and to assemble stable PSII reaction centers. These two cyt b559 mutants (H22Kα and H22YβPS+) and their PSII reaction centers will be very suitable for further biophysical and biochemical studies of the functional role(s) of cyt b559 in PSII.




Structural analysis of the L1 and HhH DNA binding motifs in the Archaeal RadA homologous DNA strand exchange protein

Li-Tzu Chen1,2, Tzu-Ping Ko2, Kuei-An Lin2, Yu-Wei Chang1,2, Ting-Fang Wang 1,2 , Andrew H.-J. Wang 1,2

1Institute of Biochemical Science, National Taiwan University, Taipei 106, Taiwan; 2 Institute of Biological Chemistry, Academia Sinica, Taipei 115, Taiwan


Archaeal RadA, like other members of RecA protein family, engages in an ATP-dependent DNA strand-exchange reaction that is performed by the ssDNA nucleoprotein helical filament and a homologous dsDNA sequence. Here we present a crystal structure of :Sulfolobus solfataricus (Sso) RadA right-handed filament at 1.93Å resolution. Compared to other known RecA family protein structures, this new crystal structure revealed a better atomic resolution at two putative DNA binding sites: L1 motif and Helix-hairpin-Helix motif. Biochemistry and genetic analyses indicated that several evolutionarily conserved structural elements in these two motifs were crucial for DNA binding :in vitro and for DNA damage repair :in vivo. A molecular model was proposed for homologous search and pairing between dsDNA and ssDNA mediated by these two DNA binding motifs.




Elucidation of the structure of a bi-cadmium compound [Cd2 (bis(2-pyridyl)formamine)3]NO3 by Cadmium-111, Cadmium-113 and Hydrogen-1 Nuclear Magnetic Resonance Spectroscopy

Hung-Chieh Lu (盧弘捷) ,Lou-sing Kan*(甘魯生)

Institute of Chemistry, Academia Sinica


[Cd2(bis(2-pyridyl)formamine)3]NO3(1) (Scheme 1) was synthesized by reaction of Cd(NO3)2with bis(2-pyridyl)formamidine in methanol. We find that Cd (111Cd and 113Cd) and 1H NMR solution studies revealed the three bis(2-pyridyl)formamine motifs of 1 not only have a three-fold symmetry about the Cd-Cd axis, but also the two pyrimidines of each of the bis(2-pyridyl)formamine moiety are symmetrical and are mirror images with respect to the plane defined by the three Cas (Scheme 1). This report presents unequivocal evidence that 1 exhibits a highly symmetric structure in solution(illustrated at bolow). Mr. Casper Wu of Rezwave Technology Inc. is appreciated for his generosity to provide [Cd2(bis(2-pyridyl)formamidine)3]NO3sample. Appreciations are also due to the High-Field Biomacromolecular NMR Core Facility, which is supported by the National Research Program for Genomic Medicine and Academia Sinica, Taiwan, ROC, for assistance of the NMR observations.




The study of DNA triplex with tight turn structure

Ming-Tsai Wey 1,2 and Lou-Sing Kan 2

1 Institute of Bioinformatics and Structural Biology, National Tsing Hua University;
2 Institute of Chemistry, Academia Sinica


We reported the structure of‘paperclip’type DNA triplex 5’-T1C2T3C4 T5C6C7T8C9T10C11T12A13 G14A15G16A17G18 that was studied by UV, CD, NMR spectroscopy and molecular dynamics simulations. Previous published structure showed the C6C7 /G18 apex base triad (in which C6C7 forms a tight turn structure or folding structure without loop) is required for intramolecular triplex formation (Biophys. J., 82, 3170 (2002)). In this study, we investigated the structure of DNA triplex that contains also a C6C7/G18 apex base triad formed by 5’-T1C2T3C4T5C6C7 T8C9T10C11T12-3’(PY) and 5’-A13G14 A15G16A17G18-3’(TFO). The molecular size of the triplex was verified by gel electrophoresis and the structure by NMR spectroscopy and molecular dynamics simulation. The structure constructed by TFO via Watson-Crick pairing to the C7-T12 and Hoogsteen base pairing to T1-C 6 of the PY. In other words, a triplex with C6C7 tight turn, similar to that of paperclip triplex, was formed. There are six base triads including the T1/T12A13 that was not observed in the paperclip triplex. However, the conformation C6C7/G18 base triad of PY + TFO is similar to that in the paperclip triplex. Therefore, the growing number of triplex formation may prove useful as therapeutic agents. (Supported by National Science Council and Academia Sinica)




Structure of a Tellurite Resistant Protein (TerB) from Klebsiella pneumoniae

Sheng-Kuo Chiang, Yuan-Chou Lou, and Chinpan Chen

Institute of Biomedical Science, Academia Sinica, Taipei, Taiwan.


Klebsiella pneumoniae (KP) is one of the top five pathogenic bacteria and causes community and hospital infection of pneumonia in Taiwan. In some immuno-deficient patients such as diabetic, KP forms opportunistic infection and causes septicemia combining with Pyogenic liver abscess. In bacteria, tellurite resistance is conferred by the ter (TeR) operon (terZABCDEF), which also controls the resistance to infection by various bacteriophages, and the resistance to pore-forming colicins. Using heteronuclear multi-dimensional NMR experiments and computational calculation, we have determined the NMR structure of TerB protein (151 residues) from KP, which is mainly composed of 7 -helices, on the basis of 2593 restraints. The ensemble of 20 NMR structures was well defined, with RMSD 0.38b0.07 Å for backbone atoms, and 1.22b0.09 Å for heavy atoms, respectively. A unique property of TerB structure is that the charged residues obviously distributed into two regions, positively charged residues locating at N-terminus, and the negatively charged residues by contrast locating at C-terminus. To the best of our knowledge, there is no any tellurite resistance protein structure reported to date, so it is likely that our structure will be the first reported structure of a tellurite resistance protein. We further performed structural similarity search using DALI method, and found that four proteins with distinct functions (cell adhesion, transferase, electron transport, and kinase) have Z-score in between 3.2 to 3.5. The detailed structural characterizations of TerB from KP will be discussed in this post presentation.




Structural and functional study of C. elegans Cell-Death Related Nuclease 4 (CRN-4)

Yu-Yuan Hsiao (蕭育源)1,2 and Hanna S. Yuan (袁小琀)2

1Institute of Bioinformatics and Structure Biology, National Tsing Hua University 2Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan


Programmed cell death (apoptosis) is an essential biological process for development, maintenance of appropriate cell number and defense against virus infection. One of the major biochemical features of apoptosis is chromosome DNA fragmentation. Cell death related nuclease 4 (CRN-4) is one of the newly identified apoptotic nucleases in C. elegans involved in the sequential DNA degradation process. CRN-4 interacts with several nucleases, including CPS-6, NUC-1, CRN-1 and CRN-5, to form a multi-protein complex, named degradeosome, which promotes apoptotic DNA degradation. A DEDDh exonuclease domain is identified in the CRN-4 and therefore CRN-4 likely bears 3’-5’ exonuclease activity, responsible for chromosome DNA nicking and/or exonuclease digestion. To reveal the molecular basis for DNA fragmentation in apoptosis, we over expressed and purified recombinant CRN-4 from E. coli. We found that CRN-4 contains both DNase and RNase activities. The crystal structure of CRN-4 was determined at a resolution of 2.5 Å by MAD method (data collected at BL12B1 in NSRRC). CRN-4 folds into a dimer with an N-terminal DEDDh exonuclease domain and a C-terminal domain with unknown function. The N-terminal catalytic domain adopts an α/β globular fold and shares similar structure and catalytic residues to those of other members of DEDDh exonucleases, such as 3’hExo and ε186. The C-terminal domain contains a novel mixed α/β fold with 7 basic residues exposing to the surface. We suggest that the C-terminal domain in CRN-4 is involved in DNA binding in a way similar to the RNA-binding SAP domain of 3’hExo. The structural comparison of dimeric CRN-4 to several nucleases further reveals that this protein likely recognizes DNA in a unique way than we had seen before in other DEDD family of nucleases.




Implementation of Projection Reconstruction-3D Triple Resonance Experiments on Bruker NMR Spectrometers

Wen-Jin Wu and Tai-Huang Huang

Institute of Biomedical Sciences, Academia Sinica


The data collection time for a typical three-dimensional triple resonance experiment can take up several days of NMR spectrometer time. One approach to shorten the data collection time is the projection reconstruction NMR (abbreviated as “PR-NMR”) method, proposed by Kupče and Freeman (Kupče et al., J. Am. Chem. Soc., 2004, v126, 6429-6440). In this approach, only several 2D sub-spectra are required in order to reconstruct a full target 3D spectrum. Typically two orthogonal 2D projections (the 2D H-N, and the 2D H-C planes) are acquired with the evolution parameter t1 or t2 set to zero. These projections define a three-dimensional lattice; all cross peaks must lie on this lattice but not all lattice points are occupied. A third 2D experiment with t1 and t2 incremented simultaneously and in a fixed ratio, generates a projection onto a tilted plane and establishes the position of all the cross-peaks unambiguously. In this poster, we present the test results of the projection reconstruction-NMR for the SH3 domain protein (62 a.a., 7.3 kDa), and a 20 kDa protein polymerase X on Bruker spectrometers. We have established a suit of 3D triple resonance experiments for sequential backbone resonance experiments, including HNCO, HN(CO)CA, HNCA, CBCA(CO)NH, HNCACB, HN(CA)CB.




Solution structure and DNA-binding properties of the three Cys2His2 zinc finger domains of Tzfp

Chun-Chi Chou1,2 Yuan-Chao Lou1 Chinpan Chen1

1Institute of Biomedical Sciences, Academia Sinica, Taipei 115, Taiwan
2Graduate insitute of life sciences,National defense Medical center, Taipei 114, Taiwan


The testis zinc finger protein (Tzfp) binds to the upstream flanking sequence of the Aie1(aurora-C). Previous studies showed that Aurora-C may play a role in centrosome function and key regulator in cell division. The mouse Tzfp gene contains a N-terminal BTB/POZ domain and three C-terminal C2H2 zinc fingers. The zinc finger domain of Tzfp (TZD) binds to the TGTACAGTGT at the Aie1 promoter with high specificity. The interaction between Tzfp and Aie1 promoter may play a important role in the biological function. To gain insight into the transcriptional mechanism, we have applied a variety of experiments for structural studies. According to circular dichroism results, the TZD is coordinated by zinc and the secondary structure is significant. Without zinc or treating with EDTA, the TZD is unfolded. By using heteronuclear multidimensional NMR experiments and structure calculation, the solution structure of TZD was determined, which contained a conserved ββα structure with very flexible orientations. Besides, the TZD was complexed with different lengths of synthesized DNA, which all contain a core motif (GTACAGT). The TZD was also complexed with the 16-mer mutational DNA, 16m5FR, which contain a point mutation. In gel shift assay, the 16-mer DNA-TZD complex obviously showed band shift but the mutant did not. By comparing the HSQC spectra between free TZD and it’s complex, we found that many residues in 12-mer and 16-mer DNA complexes show chemical-shift changes which are located in all regions of TZD, suggesting that TZD wrap duplex DNA completely. The resonances assignment of DNA-TZD complexes is in progress. In this poster presentation the structure of free TZD as well as the binding interaction in the complex will be discussed.




Crystal structure of the N-terminal domain of DnaD suggests a mechanism for DNA remodeling and DNA replication initiation

Cheng-Yang Huang1, Wei-Ti Chen1,2, Chen-Yi Tsao1, Yi-Wei Chang1, Chwan-Deng Hsiao1,2

1Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan;
2Taiwan International Graduate Program, Academia Sinica, Taipei, Taiwan


DnaD is a primosomal protein required for DNA replication of the Gram-positive bacteria. It interacts with the replication initiator protein DnaA and the replicative ring helicase DnaC. DnaD also has DNA remodeling activity to untwist supercoiled DNA via forming higher order nucleoprotein structures. Here, we report the crystal structure of the N-terminal domain of DnaD of Geobacillus kaustophilus HTA426 at 2.3Å resolution, revealing a dimer. The size exclusion chromatographic study and the analytic ultracentrifugation sedimentation velocity experiments demonstrated that the major conformation of the N-terminal domain of DnaD is dimer. However, other multimeric states are also observed following the protein concentrations increased. Based on both the structural and biochemical studies, we propose that the potential multimer formation surface and several other regions of DnaD may participate in protein-protein interaction during DNA replication. These findings may provide a structural insight into how the DNA can be remodeled by the N-terminal domain of DnaD during DNA replication initiation.




Dbv29 that is characterized as a Flavin-containing hexose oxidase completes the N-acyl-aminoglucuronyl substituent modification for antibiotic A40926

Yi-Shan Li1,2, Jin-Yuan Ho3,4, Yu-Ting Huang1,2, Chun-Yen Lee1,2, Chia-Chi Huang3,4, Syue-Yi Lyu1,2, Chuan-Jiuan Huang1,2, and Tsung-Lin Li1

1Institute of Genomics Research Center, Academia Sinica, Taipei 11529, Taiwan
2Department of Food Science, National Taiwan Ocean University, Keelung, Taiwan
3Institute of Biological Chemistry, Academia Sinica, Taipei 11529, Taiwan
4Taiwan International Graduate Program, Academia Sinica, Taipei 11529, Taiwan


With the emergence of vancomycin-resistant enterococci (VRE) and the more serious vancomycin-resistant Staphylococcus aureus (VRSA), the development of novel glycopeptide antibiotics with higher efficacy against the resistant strains is thus highly demanded. The approach through the genetic manipulation of secondary-metabolite biosynthetic pathways toward new drug development has been proven to be promising and indeed a hot research area. With gene analysis and biochemical evidence, three gene products were identified to be involved in the synthesis of N-acyl-glucosaminyl moiety. In this study, Dbv29 was further characterized as a flavin-containing oxidase that carries out a 4-electron oxidation reaction. Through assaying against a variety of enzymatically synthesized substrates, Dbv29 was concluded to complete the last step modification toward N-acyl-aminoglucuronic acid in this biosynthetic sequel. Meanwhile, molecular oxygen was determined to be required and hydrogen peroxide is a concomitant product when the reaction takes place. Taken together with gene analysis, mutagenesis and stable isotope labeling study we were able to provide insights into the catalytic mechanism for Dbv29. To our knowledge, Dbv29 may be the second characterized example for its kind. In summary, Dbv29 is a novel FAD-containing hexose glycoside oxidase that specifically catalyzes the oxidation of the C-6 primary alcohol functional group of the sugar moiety of the glycoside into the carboxylic acid counterpart. In addition, Dbv29 is found to be able to transform teicoplanin, one of the only two clinically licensed glycopeptide antibiotics, into a novel acid counterpart that may therein possess a higher biological activity and provide another front line against the bacterial resistance.




Structure-based Mechanistic Study for Three Homologous proteins, Deacetylases and Nucleotidyltransferase from the Biosynthetic Pathways of Three Clinically Important Glycopeptide Antibiotics

Hsiu-chien Chan 2,Yu-Ting Huang5, Chia-Cheng Chou1,Jin-Yuan Ho3,4, Chang-Jer Wu5, Yi-Shan Li5, Ming-Daw Tsai1,3, and Tsung-Lin Li1

1Institute of Genomics Research Center, Academia Sinica, Taipei 11529, Taiwan; 2Institute of Bioinformatics and Structure Biology, National Tsing Hua University, Hsinchu 300, Taiwan; 3Institute of Biological Chemistry, Academia Sinica, Taipei 11529, Taiwan; 4International Graduate Program, Academia Sinica, Taipei 11529, Taiwan; 5Department of Food Science, National Taiwan Ocean University, Keelung, Taiwan


With the emergence of vancomycin-resistant enterococci (VRE) and the more serious vancomycin-resistant Staphylococcus aureus (VRSA), the development of novel glycopeptides antibiotics with higher efficacy against resistant strains is thus highly demanded. The approach through the genetic manipulation of biosynthetic pathways has been proven to be promising and become a hot research area. There has recently been considerable progress in elucidating the biosynthetic steps of glycopeptide antibiotics. However, many important questions remain outstanding. In this study, three homologous (>65% identical) proteins Orf2*, Dbv21 and Orf15, previously assigned as hypothetical proteins in the biosynthesis of teicoplanin, A40926, and chloroeremomycin, respectively, are biochemically characterized as novel deacetylases (Orf2* and Dbv21) and thymidyltransferase (Orf15). On the other hand, the native structures for Orf2* and Dbv21 were determined by x-ray crystallography using MAD and MR with resolutions at 1.7 and 2.8Å, respectively. Both structures reveal an α/β fold in which a seven-stranded parallel β-sheet is wrapped up by helices. A metal-binding site that is found associating with a sequence motif HxDDx is positioned at the bottom of the putative active site. The metal ion that was chemically ripped off resulted in inactivity. Pivotal residues that surround the catalytic center have also been determined by mutagensis study. Taken together, the catalytic center and mechanism can be featured in thorough. We would thereby be in a position to further expand the substrate dimensions based on the obtained structural and biochemical knowledge so that novel compounds with higher efficacy may be invented.




Clarification of Potential Structural Element of Vigna radiate Plant Defensin 1Involving in Inhibiting Insect α-Amylase

Kuo-Chang Cheng, Tian-Ren Lee, Chao-Sheng Cheng, Ping-chiang Lyu

Institue of Bioinformatics and structural Biology, Department of Life Science, National Tsing Hua University, Hsinchu, Taiwan


Vigna radiate plant defensin 1 (VrD1) is the first reported plant defensin exhibiting insecticidal activity against Callosobruchus chinensis (bruchid). Three dimensional structure of VrD1 has been determined by nuclear magnetic resonance spectroscopy in our lab. Nevertheless, the mechanism of insecticidal activity is unclear. According to our previous work, VrD1 has shown to inhibit Tenebrio molitorα -amylase (TMA) in vitro which may trigger insecticidal activity. Computational docking model of VrD1-TMA complex also implied that loop L3 of VrD1 is important for this inhibition. Among plant defensins of known structure, VrD1 is the first case containing a 310 helix. Based on protein sequence alignments, VrD1 is different from other defensins and contains an arginine in place of glutamate at the residue 26. We propose that it may induce a shift in the orientation of Trp10, thereby facilitating the 310 helix formation and then contributing to the stability of global structure. Therefore, we focused on the influence of loop L3 and 310helix of VrD1 on its inhibitory function in this study. Site-directed mutagenesis was carried out to study critical residues of VrD1 involving in TMA inhibitory activity. The secondary structures of purified proteins were checked by circular dichroism (CD). Our results showed that several mutants significantly decreased in TMA inhibition, indicating that the loop L3 and 310 helix indeed play important roles in VrD1 insecticidal function.




Cloning, expression, purification, and crystallization of a glutaminyl cyclase from Xanthomonas campestris: the prototype of prokaryotic glutaminyl cyclase

Yu-Ruei Wang 1,2, Kai-Fa Huang 2, Andrew H.-J. Wang 2*

1Department of Biochemical Science and Technology, College of Life Science, National Taiwan University, Taipei 106, Taiwan;
2 Institute of Biological Chemistry, Academia Sinica, Taipei 115, Taiwan


Glutaminyl cyclases (QCs) catalyze the N-terminal pyroglutamate (pGlu) formation of numerous bioactive peptides and proteins from their glutamine or glutamate precursor. QCs can be classified into two sequence and structurally distinct groups: the group I is found as mammalian enzymes; the group II is found in plants, parasites and bacteria. The bacterial QCs have never been studied so far. Here, we present the cloning of a QC gene from the genomic DNA of Xanthomonas campestris, which are known as pathogen of plants. The protein was over-expressed in Escherichia coli system using a designed histidine-tag for further purification by Ni-affinity chromatography. Meanwhile, the identity of the protein was confirmed by SDS-polyacrylamide gel electrophoresis and an enzyme-coupled assay of QC activity. We have prepared more than ten site-directed and alanine-substituted mutants for the putative active-site residues. By enzyme kinetic and X-ray crystallographic studies, these mutants may provide great insights into the substrate binding and catalysis mechanism of prokaryotic QCs. In addition, we are also constructing other mutants of bacterial QC, aimed to investigate the unusual resistance of group II QCs to heat, extreme pH values, and detergents.




Crystal Structure of Laminaripentaose-Producing β-1,3-Glucanase (LPHase) from Streptomyces matensis DIC-108

Ming-Tsung Hsu1, Sheng-Wen Liu2, Yaw-Kuen Li2, Wen-Ching Wang1

1Institute of Molecular and Cellular Biology, National Tsing Hua University, Hsinchu 300, Taiwan
2Department of Applied Chemistry, National Chiao Tung University, Hsinchu 300, Taiwan


The fungal cell wall, which is mainly composed of β-1,3-glucan and chitin, is essential for the fungus but absent from the mammalian host. Thus, fungal cell wall may be an attractive target for new antifungal agents. The laminaripentaose-producing β-1,3-glucanase (LPHase) secreted from Streptomyces matensis DIC-108 belongs to a hydrolase that can break the glycosidic bonds of β-1,3-glucans into the pentasaccharide laminaripentaose. Based on amino acid sequence similarities, LPHase is assigned to a member of the glycoside hydrolase family GH-64. This investigation is aimed to determine the crystal structure of LPHase as the first step toward its detailed structure-function studies. The native protein crystal is grown by the hanging-drop vapor diffusion method about three to four days in 20℃, and the seleno-protein crystal is grown in 4℃ for least one week. The native crystal belongs to the space group P212121, with one molecule in the asymmetric unit and unit-cell parameters a = 46.16 Å, b = 60.68 Å, c = 149.40 Å, α = β = γ = 90.00°. A complete native data set was collected to 1.62 Å resolution at the National Synchrotron Radiation Research Center (NSRRC) beamline BL13C1. The phase was solved to 2.3 Å resolution using multiwavelength anomalous dispersion (MAD) data from the SeMet-LPHase crystal collected at BL12B2 Taiwan beamline at SPring-8, Japan. The electron density is continuous from Val3 to Pro279 and Val283 to Asp367. Residues that are not included in the final model are from the N-terminal segment (Met1, Ala2) and a loop region (Asn280, Asp281, and Gln282). The structure was refined to 1.62 Å resolution, to an R-factor of 25.2% (Rfree = 28.0%). LPHase structure consists of an up-and-down β barrel and mixed α/β structure. Given the first structure reported for the β-1, 3-glucan for the GH-64 family, further structure and enzymatic characterizations will be conducted to understand its limited action on β-(1,3)-D-glucans as well as the detailed catalytic mechanism for future inhibitor design.




Equilibrium folding and aggregation of wild type and familial mutants of amyloid beta in Alzheimer's disease

Hoi-Ping Shi 1, Yun-Ru (Ruby) Chen 1

1 The Genomics Research Center, Academia Sinica No. 128, Academia Road Section 2, Nan-Kang, Taipei, 115, Taiwan


The amyloidogenic protein, amyloid β peptide (Aβ) composed of 40 or 42 amino acids is a critical component in the etiology of the neurodegenerative Alzheimer's disease. Aβ is prone to aggregate and forms amyloid fibrils progressively both in vitro and in vivo. To understand the process of amyloidogenesis, it is pivotal to examine the initial stages of the folding process. We examined the equilibrium folding properties, assembly states and stabilities of the early folding stages of Aβ40 and 42. We found that Aβ40 and 42 have different conformations and assembly states upon refolding from their unfolded ensembles. Aβ40 is predominantly an unstable and collapsed monomeric species, whereas, Aβ42 populates a stable structured trimer/tetrameric species at concentrations above approximately 12.5 μM. Thermodynamic analysis showed that the free energy of Aβ40 monomer and 42 timer/tetramer are ~1.1 and ~15/22 kcal/mol, respectively while adopting two-state mechanism. A three-state model for Aβ42 with the presence of monomer is also suggested and may explain the related fibrillization patterns. Preliminary studies on familiar Aβ40 mutants including Artic (E22G), Dutch (E22Q), and Flemish (A21G) show that they are more stable than the wild type. The Artic mutant seems to adopt similar properties to Aβ42 although a peptide with 40 residues was used. Our results show that the amyloidogenic folded structure of Aβ, although considered as natively unfolded protein, is important for the formation of spherical β oligomeric species. However, Aβ oligomers are not an obligatory intermediate in the process of fibril formation because oligomerization is inhibited at concentrations of urea that have no effect on fibril formation. The distinct initial folding properties of Aβ40/42 and the mutants may play an important role in the higher aggregation potential and pathological significance.




The Effects of Macromolecular Crowding on the Conformation and Stability of Protein

Tse-Yu Chung1, Irene Deli1, Shao-Rong Luo1, Chia-Lin Chyan1

1Department of Chemistry, National Dong Hwa University, Hualien 974, Taiwan, ROC


Physiological fluid media contain macromolecules occupying a significant fraction (typically 20-30%) of the total volume. Biological macromolecules have evolved to function inside such crowded environments. It has been shown the natural and synthetic macromolecules can be used to mimic the crowding environments in cells. In this study, we applied NMR and other biophysical techniques to investigate the effect of crowding on the stability and conformation of ubiquitin and its mutants. Ubiquitin (Ub), a small protein with 76 residues, contained 5-strandedβ-sheet and one α-helix. We have cloned and overexpressed wild type ubiquitin and its mutants, F4A/F45W, V26A/F45W, and I30A/F45W in E. coli. The wild type ubiquitin was in the folded state in the pH ranged from 2 to 10, however, these three mutants were unfolded at pH below 3. We found that the existence of dextran did not change the secondary structure content when ubiquitin was in the folded state. However, the crowding condition did induce a significant amount of secondary structure in partially unfolded ubiquitin. We also found that addition of anion (Cl-) to partially unfolded ubiquitin can drive the equilibrium from the unfolded state toward the native or native-like state. The NMR structures of F4A/F45W at neutral pH and at pH 2 in the presence of high concentration of Cl- were closely identical to the native structure of wild type ubiquitin. The tertiary structure of F4A/F45W at pH 2 in the presence of crowding agents was similar to the tertiary structure of F4A/F45W at pH 2 in the presence of high concentration of Cl-.




Structural Basis of the Low Fidelity of ASFV DNA Polymerase X and Identification of a Key Fidelity-Controlling Residue

Mei-I Su1,4, Wen-Jin Wu3, Sandeep Kumar4, Ming-Daw Tsai1,2,4,5

1Genomics Research Center, 2Institute of Biological Chemistry, and 3Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan 115. 4Department of Chemistry and 5Department of Biochemistry, The Ohio State University, Columbus, Ohio 43210, USA


The structural basis of low fidelity for error-prone DNA polymerases is emerging as a subject of extensive research. DNA polymerase X (Pol X) from African Swine Fever Virus (ASFV) is the smallest known DNA polymerase (20 kDa) and also one of the most error prone, showing comparable efficiencies for G:C and G:G incorporations. So far, no structural information is available except the solution structure of free Pol X. Here we report the solution structure of ASFV Pol X in ternary complex with DNA and dGTP (opposite to a templating G), which shows unique binding modes for DNA (interacting with helix E at the fingers subdomain) and dGTP (in syn conformation forming a Hoogsteen base pair with the templating G). The latter is further supported by NMR and kinetic analyses using nucleotide analog 7-deaza-dGTP. We also identify a novel role of His115 in controlling base pairing and its mutation to Ala leads to substantially improved fidelity towards G:G mismatch by attenuating relative binding affinities of dCTP vs. dGTP. The results demonstrate that Pol X uses specific interactions and Hoogsteen base pairing to achieve low fidelity and G:G specificity.




Reconstitution of fluorescent allophycocyanin alpha-subunits and identification of residues critical for the autocatalytic bilin attachment

I-Chen Hu, Hui-Fen Lin, Tian-Ren Lee, and Ping-Chiang Lyu

Department of Life Sciences and Institute of Bioinformatics and Structural Biology, National Tsing Hua University, Hsinchu, Taiwan


Phycobiliproteins are valuable as fluorescent probes, because they are covalently attached with chromophores, giving them distinct absorption and emission spectra in the visible rang. C-phycocyanin (CPC), allophycocyanin (APC), and phycoerythrocyanin (PEC) are the three most common phycobiliproteins in cyanobacteria, and each contains the chromophore, phycocyanobilin (PCB). Chromophore lyases are necessary to catalyze the chromophore attachment to phycobiliproteins, such as CPC and PEC. However, no APC-specific chromophore lyase has been reported. In our study, chromophorylated APC α subunit (chromo-ApcA) was obtained via a spontaneous PCB attachment reaction. The extent of PCB attachment to apo-ApcA was comparable to that of the lyase-catalyzed reactions for other phycobiliproteins. These results indicate that ApcA has autocatalytic lyase activity. We further identified ApcA residues critical for producing functional protein using an alanine scan. The mutants ApcA-C81A and ApcA-C81S were non-functional, indicating that PCB was linked to Cys81 of ApcA. Asp84 was the most important residue tested, because ApcA-D84A completely lost the ability to form chromoprotein. Arg83 was also important for PCB attachment to ApcA, because its substitution caused a substantial decrease in chromoprotein yield. In addition, alanine substitution at either Tyr87 or Tyr88 affected the chromophorylation efficiency and resulted in blue-shift of their spectral properties. In ApcA and CPC α subunit (CpcA), the conserved residues surrounding and interacting with cysteine-bound PCB are Arg83, Asp84, Tyr87 and Tyr88. Because CpcA can not autocatalyze bilin attachment to itself, these four residues might be important for orienting PCB in the correct position during bilin attachment.




An unusual hydrogen bond network conserved in the catalytic center of animal glutaminyl cyclases is critical for catalysis of the enzyme

Kai-Fa Huang1,2, Yu-Ruei Wang3, En-Cheng Chang1, Tsung-Lin Chou4, and Andrew H.-J. Wang1,2,3,4

1Institute of Biological Chemistry, Academia Sinica, Taipei 115, Taiwan; 2National Core Facility of High-Throughput Protein Crystallography, Academia Sinica, Taipei 115, Taiwan; 3Department of Biochemical Science and Technology, College of Life Science, National Taiwan University, Taipei 106, Taiwan; 4Institute of Biochemical Sciences, National Taiwan University, Taipei 106, Taiwan


Glutaminyl cyclases (QCs) catalyze the N-terminal pyroglutamate (pGlu) formation of numerous bioactive peptides. The enzymes were reported to relate to several pathological processes, such as osteoporosis, rheumatoid arthritis, and melanoma. We have previously published the crystal structure of human QC, and identified an unusual hydrogen bond network in the active site, formed by Ser160, Glu201, Asp248, Asp305, and His319, within which Glu201 and Asp248 were found to bind to substrate. This hydrogen bond network is absolutely conserved throughout the animal QCs and thus represents an unique structural feature in the QC active site. In this study, we combine systematic steady-state kinetic studies and X-ray structural analyses of eleven single-mutation human QCs to investigate the roles of this hydrogen bond network and the residues that participate in this network. Our results clearly indicate that Asp305 is indeed hydrogen-bonded to Glu201 and Asp248. The mutations E201L, D248A, D305A, and D305E, disrupting the central hydrogen bondings Glu201•••Asp305 or Asp248•••Asp305 or both of them, decreased (> 66 folds) the enzyme activity significantly. However, the mutants E201Q and D248Q, having the COOH•••CONH2 hydrogen bonds to displace the central COOH•••COOH bonds, showed extremely low activities (< 0.05%), indicating that the two central COOH•••COOH hydrogen bonds, but not COOH•••CONH2, are essential to activity. The D305A, D305N, and D305E replacements raised the Km value by 3.6-17.5 folds and, surprisingly, decreased the kcat value by 66-329 folds, suggesting that Asp305 has a major role in the catalysis process. It is unexpected that the roles of the central COOH•••COOH hydrogen bonds can be slightly replaced by COOH•••water as observed in the structures of D248A and D305A, which showed to restore 1.1%-1.5% catalytic activities (kcat = 0.055-0.074 s-1), that is in sharp contrast to E201Q, D248Q, and D305E (kcat < 0.0036 s-1). In addition, our studies of the S160A, S160G, and H319L substitutions also suggest that Ser160 and His319 help to stabilize the conformations of Asp248 and Asp305, respectively. These results likely support a charge-related process during catalysis of animal QCs.




Characterizations of Enzymes Involved in the Biosyntheses of UDP-Glucose, UDP-Glucuronate and GDP-Fucose for Virulent Capsular Polysaccharide of Klebsiella pneumoniae K1-NTUH 2044

Chun-Yen Lee1,4, Jin-Yuan Ho2,4, Ming-Daw Tsai3,4, Tsung-Lin Li4

1Department of Food Science, National Taiwan Ocean University, Keelung, Taiwan
2Taiwan International Graduate Program, Academia Sinica, Taipei 11529, Taiwan
3 Institute of Biological Chemistry, Academia Sinica, Taipei 11529, Taiwan
4Genomics Research Center, Academia Sinica, Taipei 11529, Taiwan


Klebsiella pneumoniae, an enteric Gram-negative Enterobacteriaceae, is the major causative bacterium of hospital-acquired, debilitated and immuno- compromised infections in Taiwan. The hypermucoid capsular polysaccharide (CPS) produced by K. pneumoniae NTUH2044 is believed to play a pivotal role in resistance to phagocytosis and invoke the virulence. These prompt us to further understand the pathogenicity to a molecular level in particular in the CPS biosynthesis, export and regulation, in that it might shed light for new effective treatments to the disease. Previously, a trisaccharide repeating unit of the CPS was revealed comprising sugars of glucose, glucuronic acid and fucose. Nine out of 26 genes from the K. pneumoniae NTUH 2044 CPS gene cluster were identified on the basis of gene analysis. Consequently, genes, KP3702, 3703, 3704, 3708, 3709, and 3711, were cloned, expressed and in vitro characterized as enzymes responsible for the syntheses of GDP-mannose and GDP-fucose. And, genes, KP3699, 3701 and 3726, were characterized as enzymes responsible for the syntheses of UDP-glucose, UDP-glucuronate and UDP-galactocuronate. NDP-sugars thereby synthesized are building blocks for the CPS assembly and the resulting antigenicity, wherein each step is completed by a corresponding glycosyltransferase. To our knowledge, we have presented most intact dinucleotide sugar biosynthetic pathways for CPS among the published CPS gene clusters. The results presented herein should lead us a step further to the thorough understanding for the pathogenicity of the pathogen. Moreover, the enzymes characterized might also provide potential drug targets for new inhibitor developments and new routes for dinucleotide sugar synthesis.




Role of S100A13 in the FGF-1 Non-Classical pathway

Sepuru Krishna Mohan and Chin Yu

Chemistry Department, National Tsing Hua university, Hsinchu, Taiwan


Fibroblast growth factors (FGF) are key regulators of cell proliferation, differentiation and migration1. FGFs are essential for early embryonic development, organ formation and angiogenesis2. FGF also play important roles in tumor formation, inflammation, wound healing and restenosis2. However, FGF-1 and FGF-2, lack the conventional amino-terminal signal sequence required for the secretion of proteins through the classical endoplasmic reticulum (ER)-Golgi apparatus secretary pathway4,5. Maciag et al. demonstrated that under cellular stress, FGF-1 is exported through a non-classical release pathway involving the formation of a specific copper (Cu2+)-dependent multiprotein complex5,6. The protein constituents of this complex include the FGF-1 homodimer, the Ca2+ binding protein, S100A13, and the p40 form of synaptotagmin 1 (Syt1), which participates to cross the membrane7. To understand the role of S100A13 in FGF-1 releasing pathway, we have studied the interactions of S100A13 and FGF-1 by ITC and 1H-15N HSQC titration. The calcium-bound form of S100A13 binds to FGF with moderate affinity, in the micromolar range. 1H-15N chemical shift perturbation data reveal that residues in the calcium binding loop in S100A13 contribute to the binding of FGF-1. The residues located at the N- and C-termini in FGF-1 are observed to be involved in the FGF1-S100A13 binding interface. Reference: 1. Burgess, W.H., and Maciag, T. 1989. Annu Rev Biochem 58:575-606. 2. Hemker, H.C., van Rijn, J.L., Rosing, J., van Dieijen, G., Bevers, E.M., and Zwaal, R.F. 1983. Blood Cells 9:303-317. 3. Friesel, R., and Maciag, T. 1999. Thromb Haemost 82:748-754. 4. Jaye, M., Science 233:541-545. 5. Maciag, T.,, Science 233:545-548. 6. Maciag, T.,, Proc Natl Acad Sci U S A 89:10691-10695. 7. Jackson, A.,, J Biol Chem 270:33-36.




Inhibitors for Functional NP Oligomer Formation in RDRP Complex

Chao-Jung Wu 1, Yu-Fang Shen 2, Yu-Hou Chen 1, Fu-Yang Lin 1, Sean Wu 3, Wei-Hau Chang 3, An-Suei Yang 1, Gia-Lih Lin 2, Ming-Daw Tsai 1, 4

1 Genomic Research Center, Academia Sinica, NO. 128, Academia Rd. Sec. 2, Nan-Kang, Taipei, 115, Taiwan;
2 Department of Chemistry, NO. 250, Kuo Kuang Rd., Taichung, 402, Taiwan;
3 Institute of Chemistry, Academia Sinica, NO. 128, Academia Rd. Sec. 2, Nan-Kang, Taipei, 115, Taiwan;
4 Institute of Biochemistry, Academia Sinica, NO. 128, Academia Rd. Sec. 2, Nan-Kang, Taipei, 115, Taiwan


The three longest viral RNA segments named PA, PB1, and PB2, encode for RNA-dependent RNA polymerases (RDRP), and they form a complex and function as a complex biologically. In addition, a fourth protein NP interacts with RDRP and viral RNA (vRNA) during replication. It is assumed that if NP doesn’t form oligomer, it can’t activate PB1/PB2/PA complex activity in vivo. A recent Nature paper [Qiaozhen Ye et al., Nature 444:1078, 2006] showed that NP oligomerization is mediated by its tail loop, and certain amino acid residues are important for the trimer formation, which is the basic subunit of oligomerization. Without proper tail loop, NP forms monomer instead. In our results, NP protein can be overexpressed and purified from E. coli. According to our EM observation, NP forms a ring-like structure which is similar to the oligomeric NP structure in a functional RDRP complex from eukaryotes, indicating that our protein possesses its natural abilities. The AUC analysis showed that NP itself forms homo-trimers, while its complex with RNA forms three-fold oligomers. Our preliminary data also revealed that the phage-displayed tail loop peptide binds NP proteins in vitro, and the EGFP-fusion tail loop peptide is able to diminish the RDRP activity in vivo. We are looking for the tighter binding peptides from the phage-displayed tail loop peptide library, and hopefully the RDRP activity could be further blocked by the novel tighter binding peptides.




Three Dimensional Solution Structure of Calmodulin and Calmodulin Binding Domain of Calcineurin

Deli Irene1, Chia-Yen Liu1, Chih-Ching Wang2, T.-H. Lin2, and C.-L. Chyan1

1Department of Chemistry, National Dong Hwa University, Hualien, Taiwan 2Institute of Biochemistry and Structural Biology Program, National Yang-Ming University, Shih-pai, Taipei 112, Taiwan


Calcineurin (CaN), also known as protein phosphatase 2B (PP2B), is a Ca2+/calmodulin-dependent Ser/Thr protein phosphatase. CaN participates in the Ca2+ dependent signal pathways and cellular processes in the modulation of the immune response. CaN comprising two subunits, a catalytic subunit (CaNA, 60 kDa) and a regulatory subunit (CaNB, 19 kDa). CaNA contains of an auto-inhibitory domain and CaM binding domain at C terminal region. In the presence of less than 10-7 M Ca2+, CaNA tightly binds to CaNB, but the enzyme is inactive. Upon binding to CaM, CaN undergoes a conformational rearrangement, the auto inhibitory domain is displaced and thus allows for full activity. In order to elucidate this regulatory role and to gain more insight into the interactions between CaM and CaN, we used the multidimensional heteronuclear NMR techniques to determine the structure of CaM when bound with the CaM-binding domain of CaN (CaNp). CaNp forms an α-helical structure upon binding with CaM. The RMSD of backbone and heavy atoms of twenty lowest energy structures of CaM-CaNp complex calculated by CYANA are 0.61 and 1.14, respectively. Based on the intermolecular NOEs, CaM-CaNp complex is classified as 1-14 family with anchor residues I10 and L23, which has similar binding orientation to the CaMKK peptide in the 1-16 binding mode.




A phospho-counting switch for sequential activation of a checkpoint kinase cascade

Hyun Lee1,2,*, Chunhua Yuan3,*, Andrew Hammet4, Anjali Mahajan2, Chi-Fon Chang1, Jörg Heierhorst4, Ming-Daw Tsai1,2,3,5

1Genomics Research Center, Academia Sinica, Taipei 115, Taiwan. 2Biophysics Program, 3Campus Chemical Instrument Center, Ohio State University, Columbus, Ohio, USA. 4St. Vincent’s Institute of Medical Research and Department of Medicine SVH, University of Melbourne, Melbourne, Australia. 5Departments of Chemistry and Biochemistry, Ohio State University, Columbus, Ohio, USA. *These authors contributed equally to this work.


FHA domains and SQ/TQ cluster domains (SCDs) play important roles in DNA damage signalling. The Rad53-SCD1 has dual, genetically separable functions in regulating the activation of the Rad53 kinase, as well as in the subsequent phosphorylation-dependent and FHA domain-dependent activation of Dun1 kinase by Rad53 in Saccharomyces cerevisiae. However, the molecular mechanism how a single phosphorylation site cluster can regulate the ordered, sequential activation of a kinase cascade has remained a conundrum. Here we show that the SCD1 of Rad53 serves as an FHA domain-dependent phospho-counting switch for sequential activation of the Rad53-Dun1 checkpoint kinase cascade. We found that, while Rad53-FHA1 domain binds mono- and multi-phosphorylated SCD1 with comparable affinities, the Dun1-FHA domain has ~100-fold increased affinity for di-phosphorylated as compared to mono-phosphorylated Rad53-SCD1. NMR structures of the complexes demonstrate that the specificity for di-phosphorylated SCD1 results from a second phospho-threonine binding site not present in other FHA domains. Furthermore, whereas any single threonine residue of SCD1 is sufficient for RAD53-dependent stabilisation of stalled replication forks, two adjacent threonines in the Rad53-SCD1 are required for the DUN1-dependent transcriptional induction of ribonuclease-reductase (RNR) genes following replication blocks in vivo. The results indicate that the SCD1 phosphorylation state tunes the activation of the Rad53-Dun1 kinase cascade to the strength of the checkpoint signal, with a lower threshold for Rad53 activation and urgent replication fork stabilisation, and a higher threshold for Dun1 activation to increase nucleotide supply for restart of multiple stalled forks.




Crystal Structure and Functional Studies of the Single-Stranded DNA Binding Protein from Helicobacter pylori

Kun-Wei Chan, Yi-Juan Lee and Yuh-Ju Sun

Department of Life Science, National Tsing Hua University, Hsinchu


Single-stranded DNA binding (SSB) protein plays an important role in DNA metabolism, such as DNA replication, repair and recombination. The N-terminal domain of SSB forms an oligonucleotides/oligosaccharides binding (OB)-fold, which function as single-stranded DNA (ssDNA) binding domain. The C-terminal conserved tail is supposed to participate in the protein-protein interaction. SSB protein of Helicobacter pylori (HpSSB) is encoded by the ssb gene. HpSSB was expressed and purified. Fluorescence titration experiment was used to determine the binding mode of HpSSB. The result shows that the binding size of poly(dT) is about 30 nt. The crystal structure of C-terminal truncated HpSSB protein (residues 1-134, HpSSB134) complexed with dT(pT)34 [HpSSB134-dT(pT) 34] has been determined by molecular replacement at 2.3 Å. The N-terminal domain of HpSSB fold as a typical oligonucleotides/oligosaccharides binding (OB)-fold. The ssDNA wraps around the HpSSB134 protein by the electrostatic and ring stacking interactions. Four aromatic residues, Phe37, Phe50, Phe56 and Trp84, interact with the base of ssDNA by stacking arrangement. Meanwhile, several basic residues, Arg10, Arg35, and Arg36, on the surface of HpSSB134 form a significant electrostatic path corresponding to the ssDNA binding.




Structure and function analysis of new member of cyclin family: Cyclin I

Hsueh-Liang Chu1#, Chang-You Wu1,Tzu-Hsuan Chen1,3,Yi-Chen Yeh1 , Chuan-Mei Tsai2, Yi-Liang Liu1,Tzu-Cheng Lee1, Chia-Ching Chang1,2,3*

1Bio. Sci. Tech., Natl Chiao Tung Univ. 2 Natl Nano Device Lab . 3Physics, Academia Sinica


Cyclin I, a new member of cyclin family was identified in 1995. It contains a typical cyclin box at the N-terminal and a PEST sequences at the C-terminal. And the sequences of cyclin box of cyclin I is similar to the one of cyclin G and E. Unlike other cyclins, cyclin I expressed constantly during cell cycle. This indicates the function of cyclin I may be independent on the cell cycle. In order to reveal the structure and function of human cyclin I, we have cloned, expression and refolded this protein, in vitro . By Using far-western blot analysis, we found that the cyclin I bound with p21 cip1, waf1, in vitro. It interaction suggests that cyclin I may regulate p21’s activity




Crystal Structure of a Polar Flagellin Protein, FlaG from Helicobacter pylori

Jia-Yin Tsai 1, Lun-Der Lin 1, Shao-Wen Chou1, Haimei Huang 2, Yuh-Ju Sun 1

1 Institute of Bioinformatics and Structural Biology, National Tsing Hua University, Hsinchu, 300, Taiwan;
2 Institute of Biotechnology, National Tsing Hua University, Hsinchu, 300, Taiwan


Helicobacter pylori, a gram negative bacterium, has been recognized as an important pathogen that induces chronic gastric inflammation which can progress to a variety of diseases. H. pylori exhibits a right-handed helical morphology with a bundle of flagella. The flaG operon of H. pylori consists of flaG, fliD, and fliS genes under the control of a σ28-dependent promoter. The cellular processes of flaG gene are involved in the chemotaxis and motility. H pylori flaG gene encoded a flagellin protein, HpFlaG, which was cloned, expressed, and purified. The crystal structure of the HpFlaG was determined by multiwavelength anomalous dispersion at 2.8 Å resolution. The crystal structure of HpFlaG is a α/β fold with two α-helices and antiparallel β-strands constructing a palm-like structure.




Dual Binding Sites for Translocation Catalysis by E. coli Glutathionylspermidine Synthetase

Chien-Hua Pai,1,3 Bing-Yu Chiang,1,4 Tzu-Ping Ko,1 Chia-Cheng Chou,2 Cheong-Meng Chong,1 Fang-Jiun Yen,1 Shunjun Chen,5 James K. Coward,5 Andrew H.-J. Wang*,1,2,4 and Chun-Hung Lin*,1,2,4

1 Institute of Biological Chemistry, Academia Sinica, No. 128 Academia Road Section 2, Nan-Kang, Taipei, 11529, Taiwan
2 Institute of Biochemistry, National Yang-Ming University, Taipei, Taiwan
3 Institute of Biochemical Sciences, College of Life Science, National Taiwan University, Taipei, Taiwan
4 Genomics Research Center, Academia Sinica, Taipei, Taiwan
5 Departments of Medicinal Chemistry & Chemistry, University of Michigan, Ann Arbor, MI, USA


Most organisms use glutathione to regulate intracellular thiol redox balance and protect against oxidative stress; protozoa, however, utilize trypanothione for this purpose. Trypanothione biosynthesis requires ATP-dependent conjugation of glutathione (GSH) to the two terminal amino groups of spermidine by glutathionylspermidine synthetase (GspS) and trypanothione synthetase (TryS), which are considered as drug targets. GspS catalyzes the penultimate step of the biosynthesis— amide bond formation between spermidine and the glycine carboxylate of GSH. We report herein five crystal structures of Escherichia coli GspS in complex with substrate, product or inhibitor. The C-terminal of GspS belongs to the ATP-grasp superfamily with a similar fold to the human glutathione synthetase. GSH is likely phosphorylated at one of two GSH binding sites to form an acylphosphate intermediate that then translocates to the other site for subsequent nucleophilic addition of spermidine. We also identify essential amino acids involved in the catalysis. Our results constitute the first structural information on the biochemical features of parasite homologs (including TryS) that underlie their broad specificity for polyamines.




Human securing: a p53 regulator

Chung-Yiu Wu1, Hsueh-Liang Chu1, Tzu-Hsuan cheng3,Chuan-Mei Tsai2, Chia-Ching Chang 1,3,*.

1Bio. Sci. Tech., Natl Chiao Tung Univ., 2National Nano Device Laboratories, 3Physics, Academia Sinica.


The function of securin may regulate the cell cycle progression as. It is higly expressed in tumors, such as breast carcinomas and pituitary tumors. However its structure and function remained unknown. In this study we have cloned and refolded human securin from inclusion body of E. coli. Circular dichroism spectra analysis indicated that securin might not contain unique secondary structures. Far-western analysis indicated that securin binds with p53, in vitro. This binding manner may indicate that the function of p53 is regulated by securin.




Characterization of Human Kazal-Type Serine Protease Inhibitor 2 and 4

Chun-Wei Chen, Tian-Ren Lee, Ting Chen and Ping-chiang Lyu

Institute of Bioinformatics and Structural Biology, National Tsing Hua University, Hsinchu,Taiwan


Kazal-type serine protease inhibitors (SPINKs) are a family of small proteinous inhibitors containing a conserved Kazal domain. SPINKs were thought to block protein degradation in different tissues. Among the whole family, little was known about two members, SPINK2 and SPINK4. SPINK2 is probably as a trypsin inhibitor and involves in the protection of fertility, but biological function of SPINK4 is still unclear. In this study, we focus on functional determination of two protease inhibitors. Four common serine proteases including trypsin, chymotrypsin, elastase and subtilisin A are chosen to examine their inhibition ability. According to these assays, SPINK2 exhibits inhibitory activity against these proteases, except for elastase. Especially, the 50% inhibition concentration (IC50) of SPINK2 is determined to be 0.1 μM, showing its strong inhibitory ability against trypsin, but for SPINK4, it only shows low activity against chymotrypsin and subtilisin A. In the meantime, we also used site-directed mutagenesis and trypsin inhibition assay to screen the potential reactive sites (P2 and P1’) of SPINK2. As compared with other mutants and wild type, P1’ site (His 24) mutant replaced with Glu, shows 40% loss of its trypsin inhibition, implying this site plays an important role in the function of SPINK2.




Crystal structure of Helicobacter pylori formamidase AmiF reveals a cysteine-glutamate-lysine catalytic triad

Chiu-Lien Hung1, Jia-Hsin Liu 1, Wei-Chun Chiu 1, Shao-Wei Huang 2, Jenn-Kang Hwang 2, and Wen-Ching Wang 1

1Institute of Molecular and Cellular Biology and Department of Life Science, National Tsing Hua University, Hsinchu, Taiwan;
2Institute of Bioinformatics, National Chiao Tung University, Hsinchu, Taiwan


Helicobacter pylori AmiF formamidase that hydrolyzes formamide to produce formic acid and ammonia belongs to a member of the nitrilase superfamily. The crystal structure of AmiF was solved to 1.75 Å resolution using single-wavelength anomalous dispersion methods. The structure consists of a homohexamer related by 3-fold symmetry in which each subunit has an α-β-β-α four-layer architecture characteristic of the nitrilase superfamily. One exterior a layer faces the solvent, whereas the other one associates with that of the neighbor subunit, forming a tight α-β-β-α-α-β-β-α dimer. The apo and liganded crystal structures of an inactive mutant C166S were also determined to 2.50 and 2.30 Å, respectively. These structures reveal a small formamide-binding pocket that includes C166, E60, and K133 catalytic residues, in which C166 acts as a nucleophile. Analysis of the liganded AmiF and N-carbamoyl D-amino acid amidohydrolase binding pockets reveals a common Cys-Glu-Lys triad, another conserved glutamate, and different subsets of ligand-binding residues. Molecular dynamic simulations show that the conserved triad has minimal fluctuations, catalyzing the hydrolysis of a specific nitrile or amide in the nitrilase superfamily efficiently..




The crystal structure of an aliphatic amidase AimE from Helicobacter pylori

Yu-Wen Hua, Chiu-Lien Hung, and Wen-Ching Wang

Institute of Molecular and Cellular Biology, Department of Life Science, National Tsing Hua University, Hsinchu, Taiwan


Aliphatic amidase AimE, with 34% identical to AimF paralogue in Helicobacter pylori, is indicated to be responsible for ammonia production, which may also be related to selective pressure in acidic environment. AimE is a cytoplasmic acylamide amidohydrolase ( EC ) that catalyzes the conversion of short-chain amides to corresponding ammonium and carboxylic acids and belongs to a member of the nitrilase superfamily. The crystal structure of AimE was solved to 2.38Å resolution by the molecule replacement method using the backbone of amidase from Geobacillus pallidus. The AimE structure demonstrates as a homodimer in which each subunit has a four-layer α-β-β-α architecture characteristic that highly resembles the nitrilase superfamily enzymes. AimE C-terminal tail extension is essential for interacting dimeric interfaces, forming a tight α-β-β-α-α-β-β-α dimer. The putative active-site residues are identified as C165, E59, and K133, which superimpose relatively well with those (C166, E60, and K133) of formamidase AimF. This catalytic triad is well conserved in the nitrilase superfamily.




Structural Studies of the Substrate-Binding Site on the Oxygen-Evolving Complex in Photosystem II by Light-Induced FTIR Difference Spectroscopy

Hsin-Ho Huang and Hsiu-An Chu

Institute of Plant and Microbial Biology, Academia Sinica, Taipei, 11529, Taiwan, ROC.


NH3 is a structural analog of substrate H2O and an inhibitor to the water oxidation reaction in Photosystem II (PSII), making it a valuable probe for structural properties of the substrate binding site on the OEC. In our previous work, we accomplished the first S2QA-/S1QA FTIR difference spectrum of NH3 treated PSII samples [Chu et al. 2004. Biochemistry 43, 10877-10885]. We found that NH3 induced the up-shift of the 1365 cm-1 carboxylate mode in the S2QA-/S1QA FTIR difference spectra. Our results suggested that this NH3-induced FTIR spectral change of the OEC is caused by the binding of NH3 to the Mn site on the OEC that gives rise to the altered S2 state multiline EPR signal. In this work, we found that NH3-induced spectral changes in the S2QA-/S1QA spectrum of PSII were diminished when the temperature of FTIR measurement was raised from 250K to 277K. Furthermore, we have performed low-frequency FTIR measurement to identify NH3-induced structural changes of the OEC in PSII. The implication of our results on the structural mechanism of photosynthetic water oxidation will be presented.




Crystal Structure studies of Spermidne Synthase

Chen-Hsi, Chu , Jia-Yin, Tsai, Po-Kai, Lu, Haimei, Huang, and Yuh-Ju, Sun

Institute of Bioinoformatics and Structural Biology, National Tsing Hua University


Polyamines, such as putrescine, spermidine, and spermine, are polycations that mediate cell proliferation and differentiation in most organisms. They stabilize and neutralize DNA and are essential for RNA and protein synthesis. Enzymes that participate in polyamine-biosynthesis pathway are important chemotherapeutic targets in several pathogenic organisms. Spermidine synthase (putrescine aminopropyltransferase, PAPT) catalyzes the transfer of the aminopropyl group from decarboxylated S-adenosylmethionine (dcAdoMet) to putrescine in the biosynthesis of spermidine. PAPT from Helicobacter pylori (HpPAPT) and Escherichis coli (EcPAPT) were expressed and purified. The crystal structures of HpPAPT have been determined by multiwavelength anomalous dispersion (MAD) at 2.3 Å. HpPAPT consists of an N-terminal β-stranded domain and a C-terminal Rossmann-like domain, with a binding pocket between the two domains. Meanwhile, crystal structures of EcPAPT complex with substrate (dcAdoMet) were also studied by X-ray crystallography. In addition, the crystal structures of EcPAPT and HpPAPT complex with a substrate analogue, S-(5’-Adenosyl)-L-methionine chloride (SAM), and several potential inhibitors, such as methylglyoxal bis (cyclopentylamidinohydrazone) (MGBQA), and 5’-Deoxy-5’-(methylthio) adenosine (MTA), etc., were studied.




XC5848, an ORFan protein from Xanthomonas campestris, adopts a novel variant of Sm-like motif

Ko-Hsin Chin1, Sz-Kai Ruan1, Andrew H.-J. Wang2,3 & Shan-Ho Chou1,*

1 Institute of BiochemistryNational Chung-Hsing University, Taichung, 40227, Taiwan, ROC
2Core Facility for Protein Crystallography Academia Sinica, Nankang, Taipei, Taiwan, ROC
3Institute of Biological Chemistry Academia Sinica, Nankang, Taipei, Taiwan, ROC


The XC5848 gene from a plant pathogen Xanthomonas campestris pv. campestris str. 17 (Xcc) encodes a hypothetical protein comprising 102 amino acid residues, with a molecular weight of 11945 Da. A BLAST search using XC5848 sequence against the UniProtKB/TrEMBL database ( finds orthologs only in the Xanthomonas genus. No close sequence homologue could be detected for other organisms. We have determined the XC5858 crystal structure to 1.68 Å resolution, and found that XC5848 adopts a well-conserved Sm or Lsm (Sm-like) fold widely found in eukarya, arachea, and bacteria, despite the very low sequence identities. However, considerable differences with the Sm-like motif were observed at the N-terminal and internal regions. Since critical residues responsible for RNA recognition and oligomer formation of Sm-like proteins are generally absent, XC5848 could play biological roles other than RNA metabolism. To the best of our knowledge, this is the first structural report of the novel variant of the highly conserved Sm-like motif.




The Crystal Structure of XC1258 from Xanthomonas campestris: A Putative Procaryotic Nit Protein with an Arsenic Adduct in the Active Site

Ko-Hsin Chin1, Ying-Der Tsa1, Nei-Li Jan1, Kai-Fa Huang2, Andrew H.-J. Wang2,3, & Shan-Ho Chou1,*

1 Institute of BiochemistryNational Chung-Hsing University, Taichung, 40227, Taiwan, ROC
2Core Facility for Protein Crystallography Academia Sinica, Nankang, Taipei, Taiwan, ROC
3Institute of Biological Chemistry Academia Sinica, Nankang, Taipei, Taiwan, ROC


The nitrilase superfamily proteins are involved in a wide variety of non-peptide carbon-nitrogen hydrolysis reactions, characterized by a thiol acylenzyme intermediate formed through the attack of a cyano or carbonyl carbon by a novel conserved catalytic triad of Glu-Lys-Cys, to produce important natural products such as auxin, biotin, precursors of antibiotics etc. Based on extensive sequence analysis, members of the nitrilase superfamily were classified into 6 or 13 functionally district groups. However, although sequence searching can identify polypeptides as members of the nitrilase superfamily, their annotations have been perplexing. It is necessary to re-classify these protein members preferably by structure-based methods. Currently only a few protein structures are available in this superfamily, including a NitFhit protein from Caenorhabditis elegans, two N-carbamyl-D-amino acid amidohydrolases (DCases) from Agrobacterium sp. Strain KNK712 and Agrobacterium radiobacter, and a putative CN hydrolase from Sacchromyces cerevisiae strain S288C. They are either dimeric or tetrameric α-β-β-α sandwich proteins.
We now report the crystal structure of XC1258, a putative nitrilase superfamily protein from the plant pathogen Xanthomonas campestris pv. campestris str. 17 (Xcc). Based on a multiple sequence alignment, XC1258 was found to exhibit only low sequence identities with other nitrilase superfamily proteins for which the tertiary structures have been determined. We have determined the crystal structure of XC1258 to a resolution of 1.73 Å using the two-wavelength MAD approach. Interestingly, a cacodylate or dimethylarsinic acid compound was found to situate perfectly in the active region, forming a strong arsenic adduct with the active cysteine residue. This observation allows us to propose a reaction mechanism for the nitrilase superfamily proteins and suggests that their activity could be inhibited by the dimethylarsinic compound through a sulfur-arsinic covalent bond.




The crystallization of apo-form UMP kinase from Xanthomonas campestris is significantly improved in a strong magnetic field

Jhe-Le Tu1, Ko-Hsin Chin1,2, Andrew H.-J. Wang3,4, & Shan-Ho Chou1,2,*

1 Institute of BiochemistryNational Chung-Hsing University, Taichung, 40227, Taiwan, ROC
2 National Chung Hsing University Biotechnology CenterNational Chung-Hsing University, Taichung, 40227, Taiwan, ROC
3 Institute of Biological Chemistry Academia Sinica, Nankang, Taipei, Taiwan, ROC
4 Core Facility for Protein Crystallography Academia Sinica, Nankang, Taipei, Taiwan, ROC


Bacterial UMP kinases (UMPK) are crucial enzymes that are responsible for microbial UTP biosynthesis. Interestingly, eukaryotic and prokaryotic cells use different enzymes for the UMP phosphorylation reactions. Procaryotic UMPKs are thus believed to be potential targets for antimicrobial drug development. Here, the cloning, expression, and crystallization of Se-Met substituted XC1936, a bacterial UMPK from Xanthomonas campestris pathovar campestris (Xcc) are reported. The crystallization of the apo-form Upchuck was found to be significantly improved in a strong magnetic field; the crystals were diffracted to a resolution of 2.35 Å, a dramatic improvement over the original value of 3.6 Å. Preliminary structural analyses of the apo-form XC1936 using crystals grown in a strong magnetic field clearly reveal well defined loop regions involved in substrate/analog binding that were invisible before. Crystallization in a strong magnetic field thus was found to be indispensable in determining the flexible region of the XC1936 UMPK structure.